How young is your Muscle? A Machine Learning framework for motor functional assessment with ageing by NMF based analysis of HD-sEMG signal

Objective With ageing, there are various changes in the autonomic nervous system and a simultaneous decline in the motor functional abilities of the human body. This study falls within the framework improvement of the clinical tools dedicated to the robust evaluation of motor function efficiency with ageing. Method Analysis of HD-sEMG signals recorded from 32 channels during Sit To Stand (STS) test are used for the functional assessment of body muscles. For this purpose, five primary characteristic features, iEMG, ARV, RMS, Skewness, Kurtosis , are employed for the study. A channel clustering approach is proposed based on the parameters using Non Negative Matrix Factorization (NMF). Results The NMF based clustering of the HD-sEMG channels seems to be sensitive toward modifications of the muscle activation strategy with ageing during STS test. Conclusion This manuscript provides a framework for the assessment of Motor Functional Age(MFA) of subjects having a range of chronological from 25 yrs to 75 yrs. The groups were made a decade apart and it was found that the MFA varies with the level of activeness of the muscle under study and a premature ageing is observed according to the change in activation pattern of the HD-sEMG grid.